Being a realist about propositional attitudes means believing that the
diVerences between thebeliefthat P and thedesirethat P, or between the
beliefthat Pand the beliefthat Q, are real, forming part of the fabric of the
world independently of our theories, interpretations, and systems of clas-
siWcation. In which case, we had better also believe that such properties can
benaturalised.Forthe onlyrealmind-independentpropertiesthat thereare
- which are not mere reXections of our systems of classiWcation – are those
which science may discover and describe (Armstrong, 1978). But must that
mean believing that those properties can bereducedto other terms? Notice
that the various forms of naturalised semantics which we shall be consider-
ing arereductive, attempting to say what it is for a state to meanthat Pin
terms of causal co-variance, selectional history, or functional role. Indeed,
most versions of these theories attempt to lay down necessary and suYcient
conditions for a state to meanthat Pin natural terms – the exception being
Fodor (1990) who tries to provide suYcient conditions only (a point we will
return to in sections 2 and 5 below).
We shall return to the question whethernaturalisationrequiresreduction
in section 5 (arguing that it does not). For the next three sections we shall
be exploring the strengths and weaknesses of the three main naturalisation
programmes on oVer. Each of these shares the assumption that defending
the reality of content-bearing propositional attitudes means saying what it
is (or at least, what it can be) for a belief to meanthat Pin terms which do
not themselves presuppose semantic notions.
It should be noted, before we begin, that the naturalisation issue is
orthogonal to the debate between wide and narrow content which we
discussed in chapter 6, since even narrow-content theorists insist that
(tokens of) narrow contents do have truth-conditions, and are about
things in the world; they merely claim that they are nottyped in terms of
their truth-conditions. So theaboutnessof our thoughts is something
which will have to be accounted for in any case. While most of those who
have tackled the naturalisation issue have been wide-content theorists (see
below), narrow-content theorists should be equally concerned to show that
semantic relations (reference, truth, and the like) exist as part of the
natural order of the world. For it is essential to narrow contents that they
should be the kinds of state, tokens of which generally possess semantic
properties. And since we may want to explain the success of a given action
in terms of thetruthof a thought (say), or failure in terms of falsehood,
these are not optional properties of thought-contents, even for a narrow-
content theorist.
162 Content naturalised